American Collegiate Men’s An Analysis of Injuries

Kyle Flik, MD, Stephen Lyman, PhD, and Robert G. Marx,* MD, MSc, FRCSC From the Hospital for Special Surgery, , New York

Background: Reported rates and types of ice hockey injuries have been variable. Ice hockey combines tremendous speeds with aggressive physical play and therefore has great inherent potential for injury. Purpose: To identify rates and determinants of injury in American men’s collegiate ice hockey. Study Design: Prospective cohort study. Methods: Data were collected from 8 teams in a Division I athletic conference for 1 season using an injury reporting form spe- cific for ice hockey. Results: There were a total of 113 injuries in 23 096 athlete exposures. Sixty-five percent of injuries occurred during games, although games accounted for only 23% of all exposures. The overall injury rate was 4.9 per 1000 athlete exposures (13.8 per 1000 game athlete exposures and 2.2 per 1000 practice athlete exposures). Collision with an opponent (32.8%) or the boards (18.6%) caused more than half of all injuries. Concussion (18.6%) was the most common injury, followed by knee medial col- lateral ligament sprains, acromioclavicular joint injuries, and ankle sprains. Conclusions: The risk of injury in men’s collegiate ice hockey is much greater during games than during practices. Concussions are a main cause for time lost and remain an area of major concern. Keywords: ice hockey; athlete exposures; injuries; concussions

Considered one of the fastest and most aggressive team ations, eye injuries, and dental injuries are diminishing. sports, men’s ice hockey has great potential for injury. Nonetheless, blunt trauma remains the most common Players move on sharp skates at speeds of up to 30 mph on cause of injury, followed by fatigue and overuse.4,17 a solid ice surface that is confined by rigid boards along The rate and types of ice hockey injuries at different lev- the rink’s periphery. Sticks made of wood, carbon graphite, els of play vary throughout the world.1-4,7,8,10-12,14,17,19 When or aluminum are used to propel a piece of vulcanized rub- comparing injury data from American elite hockey players ber at speeds of up to 100 mph (161 km/h).15 The goalposts, with that from their European counterparts, one must which are made of steel, add an additional hazard. In such understand the subtle differences in the North American a setting, injuries are to be expected. and European games. In general, the North American Ice hockey is played in many countries throughout the style of play is considered more aggressive and physical world and has become a popular sport for both men and than that in Europe, where many prior studies were per- women in many regions of the . At the elite formed. In addition, the surface area of American rinks level, this unique team sport is played by highly condi- (approximately 1560 m2) is considerably smaller than tioned athletes and requires a combination of strength, European rinks (1800 m2). For these reasons, it may be agility, balance, skill, and controlled aggression. reasonable to expect injury patterns to be different between Specialized equipment is needed to protect players from American and European amateur elite hockey leagues. each other, the ice, boards, goalposts, skates, pucks, and The objective in this study was to describe the injury sticks. With improved protection and required use of hel- patterns in collegiate men’s ice hockey in the United mets with facemasks, certain injuries such as facial lacer- States with respect to the distribution of injuries by body region, player position, type of exposure (game vs prac- *Address correspondence to Robert G. Marx, MD, MSc, FRCSC, tice), timing of injuries, and the associated time lost from Sports Medicine and Shoulder Service, Director, Foster Center for participation. Clinical Outcome Research, Hospital for Special Surgery, 535 East 70th Street, New York, NY 10021 (e-mail: [email protected]). No potential conflict of interest declared. MATERIALS AND METHODS The American Journal of Sports Medicine, Vol. 33, No. 2 DOI: 10.1177/0363546504267349 All data were collected prospectively during the 2001-2002 © 2005 American Orthopaedic Society for Sports Medicine season. The athletic trainers of all 12 Eastern College

183 184 Flik et al The American Journal of Sports Medicine

Athletic Association Division I Men’s Ice Hockey programs analysis was performed using SPSS for Windows version were contacted and asked to participate. Eight schools 11.0 (SPSS Science Inc, Chicago, Ill). agreed and were included. An anonymous ice hockey injury data capture form was developed to record detailed information regarding each injury (Appendix, available in RESULTS the online version of the article at www.ajsm.org/cgi/content/ There were a total of 113 injuries in 23 096 AEs for a total 33/2/183/dc1). Injury data were recorded by using this overall injury rate of 4.9 per 1000 AEs (Table 1). The game method for each qualifying injury. A pilot study was per- injury rate was 13.8 per 1000 AEs (74 injuries in 259 formed 2 years before this to evaluate the form and con- games), whereas the practice rate was 2.2 per 1000 AEs firm its ease of use. (39 injuries in 676 practices) for a risk ratio of 6.3. The certified athletic trainer on each team was respon- Therefore, game injuries were 6.3 times more common sible for entering injury data at the completion of each than practice injuries (P < .001). Forwards and defense- exposure. The injury form was mailed to the study center men had similar injury rates, whereas goalies had signifi- at the end of each month, providing the opportunity to cantly lower injury rates during games (P < .05) (Figure 1). change the initial diagnosis if further testing or review of The incidence of game injuries was slightly higher in the the injury by a team physician led to a change in diagno- first half of the season (57%) than in the second half (43%). sis. For each injury, the player’s age, height, weight, and Of the 74 game injuries, 27 occurred in the first period, 27 position were recorded. Included in the information gath- in the second period, and 20 in the third period. Fifty seven ered was whether the injury occurred during practice or a percent of injuries occurred to a player who was on the vis- game, the period within the game, and the location on the iting team, and 43% occurred to a home player. ice. In addition, it was noted if the injured body part was A collision, either with an opponent (32.8%) or the protected by equipment or had been previously injured. boards (18.6%), was the cause for more than half of all The injured player and the team’s trainer determined the injuries (Figure 2). Skates, sticks, or pucks were directly direct cause of the injury. Time lost was calculated by responsible for only 11.5% of all injuries. Eight percent of adding all consecutive practices and games that were injuries were considered overuse injuries. For only 9 missed because of the injury. Additional documentation injuries did the team trainer feel that the injury was pre- included whether the injury was due to illegal activity and ventable by better equipment, conditioning, or refereeing. whether this activity was penalized. Information was also Injuries during games were related to collisions in 69%, collected on the diagnostic procedures required, the treat- whereas practice injuries were related to collisions in only ment received, and the final diagnosis. Ultimately, the 38%. Nearly 40% of all injuries occurred along the boards. trainers were asked the question, “Do you believe the Concussion was the single most commonly sustained injury could have been prevented by better equipment, injury (18.6% overall) and was responsible for nearly one conditioning, or refereeing?” quarter of all game injuries. Of the 21 concussions recorded, An athlete exposure (AE) consisted of a single player only 4 (19%) occurred during practice. Six of the 17 game participating in a single game or practice. Exposure infor- concussions were thought to be due to illegal activity, with mation was recorded based on the at-risk population no penalty called on the play. Elbowing was the most com- determined by the average number of players at each posi- mon illegal play. It was felt that the injury could not have tion participating at practices and the number participat- been prevented in 8 cases and could have been prevented ing in each game. A specific daily attendance log was not by better equipment in 3 cases. The average time loss for kept. An injury was defined specifically as any injurious each concussion was 2.1 games and 6.9 practices (approxi- episode that led to loss of participation in the immediate mately 9 AEs total). Of the 21 concussions, forwards sus- subsequent AE, whether it was a practice or a game. The tained 16 and defensemen suffered 5. injury definition was validated on the injury form by Knee medial collateral ligament (MCL) sprains were the recording specific time loss information. second most frequent injury. Interestingly, these were all Each team’s athletic trainer was responsible for com- game related; no MCL sprains occurred during practices. pleting the injury forms and returning them to the study The injury type that led to the longest average time lost center at the end of the season. Each trainer was contacted was a syndesmotic ankle sprain (“high ankle sprain”). Five bimonthly to ensure compliance. For each team, exposure such injuries resulted in a mean of 5.4 games and 14.6 information was calculated based on number of games and practices missed. Overall, injuries affected a wide variety practices and number of players at each game or practice. of body parts. See Figure 3 for a complete breakdown of All data were reviewed for accuracy and completeness by injury by anatomical location and Figure 4 for time loss by the researchers before statistical analysis. Any unclear or major injury type. missing data were clarified with the responsible trainer and/or player at the end of the hockey season. Descriptive statistics consisted of the calculation of fre- DISCUSSION quencies and percentages. Injury rates were calculated as the number of injuries per 1000 AEs. Inferential analysis This prospective epidemiological study demonstrates 2 consisted of χ tests comparing rates within subgroups. A unique injury patterns in men’s collegiate ice hockey in the P value < .05 was considered statistically significant. All United States. Although the overall injury rate is high Vol. 33, No. 2, 2005 American Collegiate Men’s Ice Hockey Injuries 185

TABLE 1 Injury Rates for Game Versus Practice, Position, Home/Away, and by Period

Rate Ratio Variable N % AEsa Rate (95% CI)b P

All injuries 113 100.0 23 096 4.9 Game 74 65.5 5377 13.8 6.3 (4.2-9.2) <.01 Practice 39 34.5 17 719 2.2 Referent Position Goalie 7 6.2 2583 2.7 Referent Defense 37 32.7 7293 5.0 1.9 (0.8-4.2) .16 Forward 69 61.1 13 220 5.1 1.9 (0.9-4.2) .09 Game injuries Home 32 43.2 2688 11.9 Referent Away 42 56.8 2688 15.6 1.3 (0.8-2.1) .24 First period 27 36.5 1792 15.1 1.5 (0.8-2.4) .30 Second period 27 36.5 1792 15.1 1.5 (0.8-2.4) .30 Third period 20 27.0 1792 11.2 Referent

aAEs, athlete exposures. bCI, confidence interval.

16 Other 14 7% Knee/Leg Back/Spine Wrist/Hand 12 22% 9% 7% 10 All players Defense 8 Forward Injuries 6 Goalie Hip/Groin 4 9% Head 2 19% 0 Games Practices Overall Foot/Ankle Shoulder Figure 1. Injury rate by position per 1000 athlete exposures. 12% 15% Figure 3. Injuries by body part. 35.0 32.8 30.0 practices games 25.0 18.6 20.0 15 15.0

Percentage 8.0 8.0 8.0 10.0 6.2 6.0 12.5 3.5 3.5 5.0 2.7 1.8 0.0 10 t te e r n k ll a te ts k rds s e c a c nen a ru h Fa o e Ot now Pu Sk pos Sti po k amm 7.5 p b n al th Ov U te o o i th g th w th i wi i w ion n w io ion llis s ion 5 o lli llis C llis o Co o C C 2.5

0

Figure 2. Cause of injuries. ry jury MCL

A-C Inju Concussion Ankle sprain when compared to other collegiate sports (National desmosis in Collegiate Athletic Association [NCAA] Injury Syn Surveillance System data), of interest is the relative and Figure 4. Mean time loss by major injury type. absolute low number of injuries that are sustained during practices compared to during games. Practice injury rates in men’s ice hockey were the lowest of all winter sports included in the NCAA injury surveillance system data in nature of the sport during competition, which would 2001. This finding is possibly a result of the aggressive explain the higher incidence of collision-related injuries 186 Flik et al The American Journal of Sports Medicine

during games. The hockey player is well protected by play, 3 were the result of elbowing. More severe penalties equipment and therefore unlikely to be injured unless con- and/or suspensions should be considered for elbowing to tact at high speed is involved (more common in a game the head to reduce these injuries. scenario). Checking (purposeful body contact in which 1 The position of the player has been reported to be an player attempts to disrupt the progress of an opposing important factor in concussion susceptibility. Of the 21 player) and other potentially injurious acts, such as slash- reported concussions, forwards suffered 16 (76%), defense- ing with the hockey stick, are less common during prac- men 5 (24%), and goalies none. Therefore, higher concus- tices than during games. sion rates exist in forwards compared to their proportion- Prior studies that have used a retrospective design were al representation on the ice (50%), whereas the rate in less accurate in identifying a true at-risk population. The defensemen is more consistent with their proportional rep- prospective design in this study allows accurate calcula- resentation on the ice (33%). When adjusted for the num- tions of specific incidence rates. Our overall injury rate of ber of players on the ice, forwards had 2.1 times as many 4.9 per 1000 AEs is consistent with prior North American concussions as the defensemen. Although these data are studies.4,5 The practice injury rate in this study of 2.2 per similar to those of a recent Canadian study of players of 1000 AEs is similar to the recent review by Ferrara and similar caliber,6 a Finnish study recently reported that of Schurr, who examined intercollegiate ice hockey injuries 9 concussions suffered in the Finnish National League via a causal analysis and reported a rate of 2.5 per 1000 during a season, 5 were sustained by goaltenders.11 This AEs.4 Our game-associated injury rate of 13.8 per 1000 difference in concussion rates and susceptibility between AEs is also similar to 14.7 per 1000 AEs reported in their American and Finnish players may be explained in part by series. The NCAA Injury Surveillance System data from the differences in the game. The American style is tradi- the same 2001-2002 season revealed a practice injury rate tionally more physical than the European style and places of 2.0 and a game rate of 19.7. The slightly higher game the forwards at higher risk for contact injuries because of injury rates are likely attributed to the fact that the NCAA the common “dump and chase” offensive strategy. In addi- includes injuries that require sutures, even if there was no tion, American ice hockey is played on a smaller ice sur- associated time loss from the injury. face. Nonetheless, we feel that referee vigilance and strict Collisions are the main cause of injury in Division I penalization for delivering direct blows to the head would Men’s Ice Hockey. One third of all injuries in this study help reduce the number of concussions suffered by ice were caused by collision with an opponent, and nearly 20% hockey players in view of these findings. of injuries were related to collision with the boards. Stuart Rule enforcement or rule changes have been effective in and Smith17 found collisions to be the cause of 51% of the the past. Watson et al collected injury and penalty data in injuries reported in their review of American Junior A to evaluate the effectiveness of the “checking from level hockey. behind” rule and concluded that the rule change led to a Early investigations performed before the mandatory safer environment in a Canadian university league.22 They facemask and helmet rules demonstrated a high incidence found a significant decrease in injury rates to the head, of facial injury and lacerations.9,13,17,18,21 Our extremely neck, and back after the institution of the rule, designed low rate of laceration injury is partly explained by the now specifically to lower the rate of cervical spine injuries in mandatory use of helmets with facemasks in NCAA hockey hockey players. Despite a high incidence of concussion in but also by our injury definition. A laceration in our inves- our study, there were no cervical spine injuries. Rules tigation was reportable only if it caused the player to miss against checking from behind are strictly enforced in col- the subsequent game or practice. Therefore, unless a lac- lege hockey in the United States. eration was major, it was unlikely to be included. One unique injury that appears to be relatively common Stick-related injuries have been the cause of up to 14% in ice hockey is the syndesmosis injury or high ankle of injuries in other studies,10,16,17 especially studies report- sprain. The elevation provided by the hockey skate blade ing injuries from European teams, whose players often are combined with high speeds and rapid direction changes not required to wear facemask protection. We found only a while skating place the ankle at higher potential for small number of injuries (1.8%) to be caused by the stick. torque injury, often related to “catching a rut” in the ice Although this finding may be mostly related to diminished with the skate blade. Also, the stiff hockey boot provides stick-related facial lacerations secondary to facemask use, stability to the ankle but perhaps places the region directly it may also reflect the improved padding in today’s equip- proximal to the boot at higher risk. Future biomechanical ment, which may diminish other stick-related injuries testing could provide further insight into this problem and such as fractures and contusions. potentially lead to alteration in skate design. Concussion remains a serious concern in this study pop- Strengths of the study include its prospective design and ulation and accounted for nearly 1 in 5 injuries. As docu- the comprehensive hockey injury data capture form that mented previously,6 hockey has one of the highest rates of was created to provide detailed information about each concussion injuries among contact sports. Consistent with injury. Standardized forms that are used across a variety other studies of hockey injuries, concussion occurred most of sports provide little detail regarding sport-specific cir- often during games (81%) compared with practice (19%). cumstances surrounding an injury. Injury was defined as This proportion is nearly identical to that reported in an event that caused a player to miss the subsequent prac- Swedish elite hockey by Tegner and Lorentzon.20 Of the 6 tice or game.10 This standard injury definition was used concussions that were considered to be the result of illegal because inconsistent injury definitions make comparisons Vol. 33, No. 2, 2005 American Collegiate Men’s Ice Hockey Injuries 187

of epidemiological descriptions of sports injuries difficult. REFERENCES Certainly, with this definition it is possible that some legit- imate injuries (such as a deep laceration) were not counted 1. Biasca N, Simmen HP, Bartolozzi AR, et al. Review of typical ice hockey injuries: survey of the North American NHL and Hockey because the player may not have missed the subsequent Canada versus European leagues. Unfallchirurg. 1995;98:283-288. exposure. 2. Bjorkenheim JM, Syvahuoko I, Rosenberg PH. Injuries in competitive An additional potential weakness of this study might be junior ice-hockey: 1437 players followed for one season. Acta Orthop the reliance on reporting of injuries by the team trainer. In Scand. 1993;64:459-461. each case, the trainer was very interested in study partic- 3. Daly PJ, Sim FH, Simonet WT. Ice hockey injuries: a review. Sports ipation and was in contact with 1 of the authors (K.F.) on Med. 1990;10:122-131. a biweekly basis. Although concern over the accuracy of 4. Ferrara MS, Schurr KT. Intercollegiate ice hockey injuries: a causal analysis. Clin J Sport Med. 1999;9:30-33. the diagnoses made by the team trainer is legitimate, it 5. Gerberich SG, Finke R, Madden M, Priest JD, Aamoth G, Murray K. was not feasible to have a physician opinion on each injury. An epidemiological study of high school ice hockey injuries. Childs If an injury required an evaluation by an orthopaedic sur- Nerv Syst. 1987;3:59-64. geon, the final diagnosis reported to us was that diagnosis 6. Goodman D, Gaetz M, Meichenbaum D. Concussions in hockey: given by the physician. It is certainly possible that the there is cause for concern. Med Sci Sports Exerc. 2001;33:2004- trainers misdiagnosed some injuries. Finally, a single-season 2009. 7. Groger A. Ten years of ice hockey-related-injuries in the German Ice injury study, even one that includes many teams, has the Hockey Federation: a ten year prospective study/523 international potential to not represent actual risk. For example, our 0% games. Sportverletz Sportschaden. 2001;15:82-86. rate of concussions to goaltenders underestimates the true 8. Jorgensen U, Schmidt-Olsen S. The epidemiology of ice hockey rate. injuries. Br J Sports Med. 1986;20:7-9. 9. LaPrade RF, Burnett QM, Zarzour R, Moss R. The effect of the mandatory use of face masks on facial lacerations and head and CONCLUSION neck injuries in ice hockey: a prospective study. Am J Sports Med. 1995;23:773-775. There was a significant difference in the injury rates suf- 10. Lorentzon R, Wedren H, Pietila T. Incidence, nature, and causes of ice fered during games and practices. Contact among players hockey injuries: a three-year prospective study of a Swedish elite ice hockey team. Am J Sports Med. 1988;16:392-396. is the major cause of injury, and this occurs mainly during 11. Molsa J, Airaksinen O, Nasman O, Torstila I. Ice hockey injuries in game situations. Concussions and knee MCL injuries are Finland: a prospective epidemiologic study. Am J Sports Med. common and are primarily incurred during games. The rel- 1997;25:495-499. atively high rate of syndesmosis injuries is unique to ice 12. Molsa J, Kujala U, Nasman O, Lehtipuu TP, Airaksinen O. Injury pro- hockey and shows little preference for game or practice sit- file in ice hockey from the 1970s through the 1990s in Finland. Am J uations. Overall injury rates were similar for forwards and Sports Med. 2000;28:322-327. 13. Pashby T, Carson JD, Ordogh D, Johnston KM, Tator CH, Mueller FO. defensemen, but the rate of concussion injury was highest Eliminate head-checking in ice hockey. Clin J Sport Med. among forwards. Stricter enforcement of rules relating to 2001;11:211-213. illegal blows to the head may reduce the rate of this sig- 14. Pettersson M, Lorentzon R. Ice hockey injuries: a 4-year prospective nificant injury. study of a Swedish elite ice hockey team. Br J Sports Med. 1993;27:251-254. 15. Sim FH, Simonet WT, Melton LJ III, Lehn TA. Ice hockey injuries. Am ACKNOWLEDGMENT J Sports Med. 1987;15:30-40. 16. Smith AM, Stuart MJ, Wiese-Bjornstal DM. Predictors of injury in ice We thank the following people for their efforts and inter- hockey players: a multivariate, multidisciplinary approach. Am J est in this study: Debbie Belcher, ATC (Brown University); Sports Med. 1997;25:500-507. 17. Stuart MJ, Smith A. Injuries in Junior A ice hockey: a three-year Steve Chouinard, ATC (Colgate University); Jeff prospective study. Am J Sports Med. 1995;23:458-461. Frechette, ATC (Dartmouth College); John Furtado, ATC 18. Sutherland GW. Fire on ice. Am J Sports Med. 1976;4:264-269. (Princeton University); Richard Kaplan, ATC (Yale 19. Tator CH, Carson JD, Cushman R. Hockey injuries of the spine in University); Ed Kelly, ATC (Cornell University); Michael Canada, 1966-1996. CMAJ. 2000;162:787-788. Knaisch, ATC (St Lawrence University); and John 20. Tegner Y, Lorentzon R. Ice hockey injuries: incidence, nature and Stawinski, ATC (University of Vermont). causes. Br J Sports Med. 1991;25:87-89. 21. Vinger PF. Ocular injuries in hockey. Arch Ophthalmol. 1976;94:74-76. 22. Watson RC, Singer CD, Sproule JR. Checking from behind in ice hockey: a study of injury and penalty data in the University Athletic Association Hockey League. Clin J Sport Med. 1996;6:108- 111.